Research Article
Stochastic Simulation of Typhoon in Northwest Pacific Basin Based on Machine Learning
Table 1
Setting of input and output layers in the neural network.
| Prediction model | Name of the neural network | Input data | Node number of the input layer | Output data | Node number of the output layer |
| Typhoon translation speed model | ANNa1 | ci, θi, ψ, and λ | 4 | ci+1-ci | 1 | ANNa2 | lnci, θi, ψ, and λ | 4 | lnci+1-lnci | 1 | ANNa3 | lnci, θi | 2 | lnci+1-lnci | 1 |
| Storm heading model | ANNb1 | ci, θi, θi-1, ψ, and λ | 5 | θi+1-θi | 1 | ANNb2 | ci, θi | 2 | θi+1-θi | 1 |
| Typhoon intensity model | ANNc1 | pi, pi-1, and Tsi | 3 | pi+1-pi | 1 | ANNc2 | pi, Tsi | 2 | pi+1-pi | 1 |
|
|